Well, I tinkered around a bit with Speech Note which has a good amount of features and is easy to install as a Flatpak. I think it has an option to do this, but requires a bit off fiddling, an extra tool and permissions for the Flatpak. I didn't find any software with a particularly good integration into the Desktop, though.
Also read about Blahst but didn't try it yet. Maybe that one is an option.
I tried this in Fedora/Wayland previously, and it seems to work in most applications. It uses "VOSK" models which the GUI can download automatically - you just pick your language and desired model size when setting it up.
When I was exploring this a few months ago, I noticed speech recognition models have moved on quite recently (e.g. https://github.com/openai/whisper which can be run locally) but didn't see anything integrating it into an input-method like the above.
13 replies
Not sure about your specific needs, but I saw this recently:
https://handy.computer/
Requires X11, doesn't work well on wayland.
I never used this, but what an interesting question. KDE connect seems to be able to input text from voice.
https://www.kubuntuforums.net/forum/currently-supported-releases/kubuntu-22-04-jammy-jellyfish/software-support-be/664506-kdeconnect-voice-to-text-in-kde-not-working
https://youtu.be/5d_jGOGRSMc?t=1047
True, you could use text input and some voice input on the phone, like FUTO voice input
Even Gboard?
Yes of course, it inputs text.
But gboard is surveillance software, your voice is literally analyzed on their servers.
I did some network logs of gboard, you might be interested XD it basically sends data all the time
Yes of course, it inputs text.
But gboard is surveillance software, your voice is literally analyzed on their servers.
I did some network logs of gboard, you might be interested XD it basically sends data all the time
If you do not have a real-time requirement, and a bad CPU/GPU I can recommend whispercpp. https://github.com/ggerganov/whisper.cpp
It is quite fast and can transcribe with timestamps. I guess if you have a fast CPU/GPU this transcription can happen faster than real-time.
Well, I tinkered around a bit with Speech Note which has a good amount of features and is easy to install as a Flatpak. I think it has an option to do this, but requires a bit off fiddling, an extra tool and permissions for the Flatpak. I didn't find any software with a particularly good integration into the Desktop, though.
Also read about Blahst but didn't try it yet. Maybe that one is an option.
I've used Sayboard and Whisper, and liked both of them. Sayboard is faster, Whisper handles punctuation better.
I use Talon Voice
https://github.com/Manish7093/IBus-Speech-To-Text
I tried this in Fedora/Wayland previously, and it seems to work in most applications. It uses "VOSK" models which the GUI can download automatically - you just pick your language and desired model size when setting it up.
When I was exploring this a few months ago, I noticed speech recognition models have moved on quite recently (e.g. https://github.com/openai/whisper which can be run locally) but didn't see anything integrating it into an input-method like the above.
https://numenvoice.org/